22results about How to "Reduce optimization costs" patented technology

The invention relates to a vacuum test device and method for leakproofness detection of a patrol bomb launch canister. The leakproofness qualification of the launch canister is judged by putting the launch canister into a vacuum test cavity, vacuumizing and observing the pressure drop of a vacuum gauge. The device is formed by sequentially connecting an exhaust valve, a vacuum test cavity (including a vacuum meter), a control valve, a pneumatic pipeline, a pneumatic flow regulating valve, a switching valve, a vacuum buffer tank (including a stop valve) and a vacuum pump according to the air exhaust direction. During vacuumizing, the vacuum pump is started to vacuumize the vacuum buffer tank, after a certain vacuum degree is achieved, the buffer tank serves as a vacuum generator to vacuumize the vacuum test cavity, the air exhaust rate is adjusted and set through the pneumatic flow adjusting valve, and opening and closing control over an air channel is achieved through the control valve. The vacuum buffer tank serves as a vacuum generator, frequent starting and stopping of a vacuum pump are avoided, and the service life of equipment is prolonged; the test requirements of multiple products can be met through one-time evacuation, and the test efficiency is improved; the air exhaust rate is stable and adjustable, and the consistency of product test states is ensured.

The invention discloses a multipath channel parameter reconstruction and beam coverage prediction method, and relates to the technical field of wireless communication. Aiming at the problems of low data utilization efficiency, limited environment modeling precision, strong coupling of measured data and antenna beam configuration and the like in the prior art, the method comprises the steps of configuring antenna beams and orientation parameters, collecting reference signal receiving strength, performing deterministic channel modeling simulation, establishing a mapping model, performing sparse recovery, constructing a channel propagation model, evaluating a channel map and the like. In combination with a coherence perception weighting network (CARE-Net) algorithm and ray tracing (RT) physical prior guidance, accurate reconstruction from a low-dimensional reference signal receiving strength (RSRP) measurement value to a high-dimensional multipath parameter is realized. The method has the advantages that the reconstructed multi-path parameters are decoupled from the antenna configuration, the method has the cross-beam / cross-configuration generalization prediction capability, the channel coverage performance under different antenna configurations can be accurately predicted, a reliable basis is provided for wireless network optimization, and the method is low in cost, high in precision and strong in generalization.

The invention discloses a deceptive texture-based black box adversarial sample generation method, which comprises the following steps of: acquiring a plurality of initial optimization variables, the initial optimization variables being an array of 2 + m * m dimensions, the first two-dimensional numerical values being used for representing coordinates of an adversarial patch center in an image, the last m * m-dimensional numerical values being used for representing a pixel value of an adversarial patch, and the pixel value being 1 or 0; decoding the last m * m-dimensional numerical value of the initial optimization variable to obtain a corresponding adversarial patch, and pasting the adversarial patch in the image according to the first two-dimensional numerical value of the initial optimization variable to obtain an adversarial sample; and inputting the adversarial samples corresponding to the plurality of initial optimization variables into the deep neural network model to obtain a detection result output by the model, and performing optimization variable iterative updating based on the plurality of initial optimization variables and the detection result to obtain a final optimization variable and a corresponding adversarial sample thereof. The method can improve the cheating of the anti-patch, improves the attack effect of the anti-sample, improves the optimization efficiency, and reduces the optimization cost.

The utility model provides an assembly type corrugated steel web steel-concrete composite beam which comprises a prefabricated composite concrete top plate (1), a steel bottom plate (2) and corrugated steel web supporting mechanisms symmetrically arranged in a box beam, and the prefabricated composite concrete top plate (1) is provided with shear grooves (10) and cluster type stud shear keys (11). The mounting position is arranged right above the flange framework (5); the corrugated steel web supporting mechanism comprises a perforated top plate cross beam (4), a flange framework (5), a flange small longitudinal beam (6), a top plate-web connecting piece (9), a corrugated steel web (3), a profile steel inclined strut (7) and an inclined strut anchor plate (8); according to the corrugated steel web box girder, a prefabricated assembly type component is adopted, the structure is simple, construction is convenient, the weak portion of the corrugated steel web box girder is secondarily reinforced in the longitudinal direction and the transverse direction through the corrugated steel web supporting mechanism arranged in the box girder, the structural strength of the weak portion and the supporting strength of the corrugated steel web are improved, and the service life of the corrugated steel web box girder is prolonged. And generated vibration can be transversely buffered, and the safety performance is improved.

The embodiments of this specification disclose a method for training a privacy erasure model, which uses exactly the same training method to train a large model on the original training data and the desensitized data of the original training data, respectively, and records the weights of the same large model on the original training data and the weights on the desensitized data, thereby forming a new erased data training set. A privacy erasure model can be trained using the erased data training set to erase the weight data related to the privacy data in the large model. This method directly modifies the parameters of the large model by training the privacy erasure model, fundamentally deletes the privacy data in the memory of the large model, and has extremely high security. This method also does not require retraining of the large model, which greatly reduces the optimization cost of the large model. The training device, privacy erasure method and device of the privacy erasure model described in the embodiments of this specification also have the above-mentioned effects.

The utility model discloses a metal powder preparation device which comprises powder manufacturing equipment and a powder collecting tank, a feeding port in the top of the powder collecting tank is in butt joint with a powder outlet in the bottom of the powder manufacturing equipment, a plurality of butterfly valves which are distributed up and down at intervals are arranged on the powder collecting tank, and when the butterfly valves are closed, the powder outlet of the powder collecting tank is closed. Valve plates of the butterfly valves divide an inner cavity of the powder collecting tank into a plurality of powder storage units which are independent up and down. The device can be used for high-throughput powder preparation experiments, powder characteristics of multiple groups of different processes can be obtained through one experiment, and the device has important significance in researching the influence of different process parameters on the powder preparation characteristics.

The application discloses a UT transformation-based impact small celestial body orbit terminal state optimization method and belongs to the technical field of aerospace. The application realizes the method as follows: positions and velocities of the earth and small celestial bodies are obtained according to ephemeris and kinematic equations of the small celestial bodies, and the velocity variation of the small celestial body after the impact of a probe is defined as a constant, so that two angles of the velocity change direction of the small celestial body after the impact of the probe are taken as impact design variables; UT transformation is used to perform error evolution of the small celestial body, so that the space-time distribution probability of the target small celestial body in a future period of time is obtained; the space-time distribution probability of the small celestial body is calculated, the instantaneous impact probability under the given impact design is calculated based on a Legendre-Gauss quadrature method, and the overall impact probability of the small celestial body is obtained by integral based on a Newton-Cotes quadrature method; and based on the impact probability calculation method, the impact design variables are optimized in combination with a global optimization algorithm, so that the impact probability of the small celestial body under the optimized variables is the lowest.

The invention discloses an aluminum ingot turning and stacking mechanism and a stacking robot thereof, and the aluminum ingot turning and stacking mechanism comprises a mounting seat; the transfer seat is rotationally connected to the mounting seat, and a driving motor used for driving the transfer seat to rotate relative to the mounting seat is arranged on the mounting seat; the two clamping seats are connected to the transfer seat in a sliding mode, and a driving mechanism used for driving the two clamping seats to synchronously and reversely move relative to the transfer seat is arranged on the transfer seat; wherein grooves are formed in the opposite side faces between the two clamping seats, a containing channel used for guiding a plurality of trapezoidal columnar aluminum ingots to be horizontally arranged in sequence in a straight line to form an aluminum ingot lamination layer is formed, and the two clamping seats achieve the effect that the aluminum ingots in the aluminum ingot lamination layer are overturned in the vertical direction at intervals through an overturning mechanism. According to the aluminum ingot stacking device, the layer-by-layer stacking effect that aluminum ingots on the same layer are sequentially arranged in a forward and reverse mode can be achieved, and the structural stability of the aluminum ingot stacking device is improved while the space occupied by stacking of the aluminum ingots is reduced.

The invention relates to an encryption and decryption method and device, equipment, a medium, a program product and a chip, and the encryption and decryption method comprises the steps: operating a first preset algorithm and a second preset algorithm to execute an encryption process or a decryption process, and determining target output information. Through the design of the first preset algorithm and the second preset algorithm, the first operation of the first preset algorithm and the second operation of the second preset algorithm are in one-to-one correspondence with the second operation of the second preset algorithm and the first operation of the second preset algorithm within the preset time range of the running state of the first preset algorithm and the second preset algorithm. Therefore, the overall power consumption at each moment in the preset time range is the sum of the power consumption of the first operation and the power consumption of the second operation, efficient defense against side channel attacks and error injection attacks is achieved in a mode of controlling the power consumption to be constant, and the safety and stability are further improved. And the method has the characteristics of simplicity in implementation, low optimization cost, high universality and the like.

The present invention provides a method for progressive optimization design of spiral chute structural parameters, which relates to the technical field of spiral chute optimization design. The method establishes a numerical simulation model of a gas-liquid-solid three-phase flow field, verifies and optimizes the numerical simulation model to achieve a predetermined standard. Confirm the first-level optimized structural parameters, randomly generate multiple groups of first-level optimized structural parameters as sample solutions, and use the RBF neural network training model to predict the sorting efficiency and flow uniformity, and finally calculate the fitness value and perform genetic algorithm optimization. After optimization, further correction is performed to obtain the optimal parameter value as the first group of optimized structural parameters. Then, the cross-section of the spiral chute is subdivided into straight segments and curved segments to confirm the second-level optimized structural parameters, and fitness maximization is continued to be selected as the second optimization target, and the same method is used to obtain the optimal value of the second-level optimized structural parameters. The method aims to improve the sorting efficiency and flow uniformity of the spiral chute to meet engineering design requirements.

The invention relates to the field of artificial intelligence, can be applied to business system platforms of finance, medical health and the like, and discloses a generative iterative optimization method, device and equipment for health consultation and a medium, and the method comprises the steps: receiving a health consultation question submitted by a user; according to the user portrait, driving a dialogue generation model of the current version to generate a plurality of candidate answers according to a corresponding generation strategy; performing multi-dimensional evaluation and confidence coefficient calibration on the plurality of candidate answers, screening out answers to be corrected, and obtaining corresponding corrected answer samples; constructing a structured weighted preference data set according to the evaluation and confidence data of the corrected answer sample; and performing confidence-weighted dynamic fine tuning optimization on the dialogue generation model of the current version through the structured weighted preference data set to obtain a dialogue generation model of a new version. By performing multi-dimensional evaluation on the plurality of candidate answers and screening and constructing the preference data with the quality weight for dynamic fine adjustment, the optimization cost is reduced and the optimization efficiency is improved.

The invention relates to the field of avionics bus signal integrity design, in particular to a method for optimizing ARINC659 bus signal integrity. Comprising the following steps: building an ARINC659 bus signal backboard architecture model; the terminating circuit structure of the transmitting and receiving end is optimized; refining resistance parameters according to the structure of the terminating circuit of the receiving and transmitting end; refining parallel inductance parameters according to the structure of the terminating circuit of the receiving and transmitting end; refining internal direct-current resistance parameters of the parallel inductors; a backboard terminating circuit structure is optimized; and refining series inductance parameters according to the backplane terminating structure. The method can be suitable for a multi-branch circuit signal integrity optimization scene, and impedance consistency is improved.

The invention is suitable for the technical field of computer-aided structure design, and particularly relates to a concrete mixer barrel structure optimization design method which comprises the following steps: acquiring a reference object and an initial structure parameter corresponding to the reference object; performing a first detection operation on the reference object based on an initial pitch angle determined by the initial structure parameter, and obtaining an adjusted pitch angle; performing second detection operation on the reference object based on the adjusted pitch angle and the initial structure parameter, and acquiring a detection information set in real time; determining optimization data based on the detection information set; performing optimization adjustment on the initial structure parameters based on the optimization data to obtain optimized structure parameters; wherein the optimized structure parameters refer to the lead angles of the helical blades at the axial positions in the optimized reference object. According to the optimization design method for the cylinder structure of the concrete mixer truck, the mixing uniformity of the concrete mixer truck on internal concrete can be improved.

This invention belongs to the field of waste plastic pyrolysis resource utilization, specifically a machine learning-based process optimization method for preparing light aromatics from waste plastic pyrolysis. This method constructs a dataset based on raw material characteristics, pyrolysis conditions, and product data from published literature, performs preprocessing, trains and optimizes various machine learning regression models, and selects the optimal model for predicting light aromatics yield based on evaluation indicators. Then, interpretability analysis is used to identify key factors affecting the yield, and raw material structural characteristics such as saturated C-H bonds, methyl branches, and carbonyl groups are used as inputs to enhance the characterization of raw material differences. Under industrial constraints, with the light aromatics yield as the optimization objective, a particle swarm optimization algorithm is used to globally optimize the reaction condition parameters to obtain the optimal combination of operating parameters. This invention can achieve light aromatics yield prediction and operating parameter recommendation under fluctuating raw material composition, improving prediction stability and generalization ability, reducing experimental iterations, and lowering optimization costs.

The invention discloses a selective laser melting additive manufacturing process optimization method based on an Abaqus thermal field analysis model, and the method comprises the steps: determining the shape and size of a molten pool corresponding to different process parameters through selective laser melting single-layer forming, taking the shape and size of the molten pool as reference data of the model, building a'process parameter-molten pool size 'molten pool thermal field model, and using a numerical simulation method to simulate the shape and size of the molten pool; a molten pool temperature field is calculated, a proper temperature field window is found in combination with a density and forming reliability analysis result, and defect formation is predicted and avoided according to temperature field analysis, so that forming process optimization design is scientifically and efficiently carried out, and alloy selective laser melting process optimization is realized. Therefore, the problems of high optimization cost, low efficiency and the like of the current process are solved.

This invention discloses a UAV path planning method based on multi-resolution hierarchical airspace, comprising: dividing the task airspace into non-overlapping grids of different resolutions according to altitude or environmental complexity; mapping obstacle information and labeling grid attributes; abstracting free grids into nodes and constructing a graph structure of intra-layer and inter-layer connecting edges; wherein, in each inter-layer connecting edge, the parent node of the free node in the high-resolution layer and the free node in the low-resolution layer satisfy the condition for establishing an intra-layer connecting edge; searching for the optimal discrete path from the starting point to the ending point using the A* algorithm based on the graph structure; performing curve fitting and smoothing on the path point sequence, and combining UAV dynamics constraint optimization to generate a continuous and flyable 3D trajectory. This invention adapts to different airspace complexities through hierarchical grids, ensuring planning accuracy in complex areas while improving planning efficiency in large-scale airspaces, and the final trajectory balances optimality and practical flight feasibility.

The application discloses an information display method and device, electronic equipment and a storage medium, and the method comprises the following steps: sending a first cluster information acquisition instruction to a target terminal running a target application program; the target application program is an application program developed based on an Unreal Engine; acquiring first cluster information returned by the target terminal; the first cluster information comprises root object information of each object cluster corresponding to the target application program in the memory of the target terminal, and the root object information indicates the name and type of a corresponding root object; and the root object information of each object cluster is displayed. The application actively acquires and displays more effective cluster information, which is beneficial to improving the optimization efficiency of cluster-based garbage collection and reducing optimization cost.

A method for extracting triples from knowledge graphs in the mining industry based on ontology constraints and thought chain principles is proposed. The method constructs a highly complete mining domain ontology, acquires and preprocesses unstructured text data sources from the mining domain, and then constructs structured prompts containing ontology constraint information and thought chain reasoning instructions, transforming the triple extraction task into a constrained generative reasoning process. The structured prompts are input into a generative large language model. After receiving the input, the model, relying on its self-attention mechanism and pre-trained knowledge, strictly follows the pre-defined thought chain paths in the structured prompts to conduct reasoning. The output of the generative large language model is parsed using a parsing algorithm, separating the thought chain part and the final result part of the model output. Triple data structures are extracted from the final result part to form a candidate triple list, followed by ontology-based post-processing verification and optimization. This method enables efficient, accurate, and standardized extraction of knowledge in the mining domain.

The utility model provides a window glass lifter motor mounting structure and a vehicle, and relates to the technical field of vehicles, the window glass lifter motor mounting structure comprises a first connecting piece, a motor mounting seat and a vibration reduction piece, and the first connecting piece is used for being connected with a vehicle door metal plate of the vehicle; the motor mounting base is connected with the first connecting piece, and a preset gap is formed between the adjacent side faces of the motor mounting base and the automobile door metal plate. The preset gap is filled with the damping piece. According to the technical scheme, the matching difficulty can be reduced, and abnormal sound and noise can be reduced under the condition that the arrangement design of the vehicle window of an existing vehicle type and the arrangement design of related guide rails are not changed.

The invention belongs to the field of structural design of high-speed aircrafts, and discloses a multi-stage variable hybrid optimization method for a radial force thermal wing structure of an aircraft, which comprises the following steps: acquiring an original configuration of the structure to obtain an original base structure; establishing a multi-stage variable hybrid optimization model, performing multi-stage variable hybrid optimization on the original base structure by using a genetic algorithm, and obtaining an optimal structure when a target function of the multi-stage variable hybrid optimization model tends to be stable; and judging whether the obtained optimal structure meets a design target or not, if the optimal structure meets the design target, outputting a final structure, otherwise, changing constraint conditions to reconstruct the multi-stage variable hybrid optimization model to continue optimization, and outputting the final structure meeting the design target. According to the method, a bearing / heat transport integrated path is considered at the same time, optimization variables of the beam section shape and the heat transport structure size are increased, multi-stage variable hybrid optimization is completed, the bearing and heat transport capacities of an aircraft structure are improved, and the performance of the radiation beam type wing structure under typical force heat loads is improved.

The present disclosure provides a kind of MCS target code rate dynamic correction method and device, DSS base station.MCS target code rate dynamic correction method includes: whether the PRB scheduled by NR terminal is located in NR exclusive area is detected;If the PRB scheduled by NR terminal is not located in NR exclusive area, the CQI sent by NR terminal is obtained;The MCS target code rate is corrected using the obtained CQI, the position of NR terminal and LTE CRS port number, DSS scheduling bandwidth and SSS bandwidth;Link adaptation coding and modulation are carried out using the MCS target code rate after correction.The present disclosure can effectively improve the downlink rate of NR terminal, improve the performance and capacity of NR in DSS, improve DSS resource utilization, improve NR user experience, and reduce DSS optimization cost.

The invention relates to a method for optimizing spray-formed ultrahigh-strength 7-series aluminum alloy components, in particular to a method for optimizing spray-formed ultrahigh-strength 7-series aluminum alloy components through thermodynamic calculation. The problem that components of the spray-formed ultrahigh-strength aluminum alloy are difficult to optimize is solved. According to the method, the mechanical property parameters of the common second phase in the alloy can be obtained by carrying out first principle calculation on the common second phase, and screening of the strengthening phase with the highest comprehensive performance is facilitated. The calculation of the non-equilibrium solidification process of alloys with different components is used for predicting the influence rule of the non-equilibrium solidification process of spray forming on the alloy solidification structure, the calculation of the alloy phase diagram can optimize the content of each alloy element, then the optimization of the alloy components is guided, and high-efficiency and low-cost optimization is carried out on the components of the spray forming ultrahigh-strength aluminum alloy. And an efficient and cheap way is provided for optimization of the spray-formed ultrahigh-strength aluminum alloy, and application of the spray-formed ultrahigh-strength aluminum alloy to major engineering structural parts is facilitated.